Here's my theory

About pricing

[universetoday]

Like many of you, I obsess about how to make the most revenue out of Adsense. I've made dramatic changes to improve the CTR, and watched the overall EPC plumet. As traffic goes up, EPC goes down. No matter what I'm able to do, I can't seem to make headway.

I recently signed up with Adwords to see what it would cost me to buy traffic into my site, and then to offset this with Adsense revenue. Anyway, the lowest price you can pay is $.05, so I'm assuming Adsense publishers will earn some fraction of this.

Everything above is on a sliding scale. So, if you're willing to pay a lot of money per click, you can stay on top in the competitive areas. However, if you just go for that $.05, then that only really kicks in if there's excess inventory.

So my theory is that the ad revenue we're making as publishers is on a sliding scale (obviously) by keyword. And it's probably allocated on a site by site basis. Once you serve up all the high-paying ads, Google starts serving up the lower paying stuff, and finally, there's probably a bottomless well of people willing to pay $.05.

I don't think it has anything to do with Smart Pricing, or ROI for advertisers.

If you don't send a lot of visitors on a day, you'll maintain a nice high EPC, but as you boost up, you're dipping deeper and deeper into that $.05 inventory.

So, how to test this...

What I would probably do is identify the different keywords being served by my site. Although my site has a specific focus, it does bleed into different areas. Use Adsense channels to track what kinds of EPC your site is earning for different keywords (easier said than done). If I'm right, you should see that the different keywords are paying out different EPCs, and the ones that you're serving up a lot of will have that low EPC.

If that's true, then the solution is to probably diversify your site content. Figure out ways to develop content in your site which will appeal to different keyword advertisers. Don't worry if you get a lot of traffic to those pages, because the EPC should be pretty high.

Anyway... just a theory. :-)

[wyweb]

interesting...

[Alioc]

This is being done by many people already. :) However, depending on EPC will only make you happy when you check your stats and see how high it is. It is the money you earn which does matter, not EPC.

So:

Don't worry if you get a lot of traffic to those pages, because the EPC should be pretty high.

We can say that if you stop worrying about traffic, you will not make it.

[incrediBILL]

And it's probably allocated on a site by site basis.

You're somewhat correct, have you ever read the page about smart pricing?

[adwords.google.com...]

Read "How smart pricing works"

It also depends on what people BID, when I run AdWords ads I just set up $0.05 ads to run on the content network and get many thousands of impressions, go figure. That's not even smart pricing, that's just smart business - only pay if it runs for as minimum as possible.

[Alioc]

That's not even smart pricing, that's just smart business - only pay if it runs for as minimum as possible.

incrediBILL. shhht! ;)

[universetoday]

Of course we have to worry about traffic, but think this runs counter to the theory that you should remove ads from the poor performing pages. That shouldn't make a difference. If you remove ads from these pages, your overall EPC will go up, but your total income will still be a little less.

Instead, you want to work on building traffic to a set of pages until the EPC starts to dip down - in other words, you're exhausting the high-paying inventory for that keyword for your site. Then build up traffic for a different keyword on your site, or start a different site.

Or maybe you could start a different site and hive off your traffic, but I'll bet the high-paying keywords are allocated on an account by account basis, which is why we're only allowed to have a single Adsense account.

[Alioc]

I'll bet the high-paying keywords are allocated on an account by account basis

I don't bet, I win or don't play at all. :) We operate many websites which are tied a single account and I see constant high EPC on some of them and constant low EPC on others. Except this point, you're correct - I think.

[universetoday]

But are your websites focused on the same keywords or different ones? If they're different, they should be drawing from different pools of inventory, and bottom out at different levels of traffic. So, some will have higher EPCs than others based on what advertisers are paying for keywords, etc. But you should get a sense of when one of your sites is dipping down into the clearout inventory, and then focus on other sites.

[Alioc]

One very huge site with no specific topic; talks about almost anything. High traffic, low EPC. Getting better though, as the traffic gets more targeted.

Other sites cover very specific topics. As a result, return high EPC. As for the CPC of the ads on these sites, I beleive we're getting what the sites deserve individually. I can't detect any loss account-wise because of the -low EPC huge site-.

Just what I see... I think it's all about keeping the target quality of traffic as you grow. Otherwise, EPC falls down as you say. Same result but different reason. ;)

But you should get a sense of when one of your sites is dipping down into the clearout inventory, and then focus on other sites.

As I said above, I don't beleive in "clearout inventory" I call the same effect "decreasing of traffic targeting" While building new sites for more revenue, it is wise to work on old ones to keep them alive and targeted. If you do this, you won't see that effect.

However, whatever you do to keep the traffic quality as high as possible, considering the percent of converting traffic to bandwidth spenders, it is inevitable to keep things as good as they were once in a time.